Astrometry of circumstellar masers

a rXiv:as tr o-ph/28317v116Aug22ASTROMETRY OF CIRCUMSTELLAR MASERS H.J.van Langevelde Joint Institute for VLBI in Europe Postbus 2,7990AA Dwingeloo,the Netherlands langevelde@jive.nl W.H.T.Vlemmings Sterrewacht Leiden Niels Bohr weg 2,2300RA Leiden,The Netherlands vlemming@strw.leidenuniv.nl Abstract The circumstellar masers around evolved stars offer an interesting pos-sibility to measure stellar parameters through VLBI astrometry.In this paper the application of this technique is discussed,including the accu-racy and the uncertainties of the method.The different maser species (OH,H 2O,SiO)have slightly different characteristics and applications.This paper does not concern astrometry of maser spots to study the kinematics of the envelope,but concentrates on attempting to measure the motion of the underlying star.Keywords:masers –stars:circumstellar matter –techniques:interferometric –techniques:astrometryIntroductionThere are a number of scientific justifications for astrometry of stars with circumstellar masers.Historically,the registration of radio single dish and infrared data was so uncertain that proper identification of maser bearing stars was a problem.Most of these problems have been overcome;the accuracy of connected element interferometers is sufficient to make these identifications unambiguously.However,there is still a lot of uncertainty when comparing VLBI data from different experiments.Notably,when maser distributions from different transitions are over-layed,a common origin is usually assumed.Astrometry can be done to do such comparisons properly.12However,the most tantalizing application of astrometry of circum-stellar masers is the accurate determination of the motion of the star on the sky.Proper motions of a star anywhere in the Galaxy(10km/s at 8kpc corresponds toµ=0.25mas/year)can in principle be obtained with VLBI,as can the parallax for such distances(π=0.125mas).A direct distance estimate is important for studying the physics of the mass-losing stars.For many Mira variables Hipparcos measured the dis-tance,but for the most enshrouded stars VLBI offers the best approach. Both the proper motion and distance are inputs for studying the stellar population and Galactic kinematics(Habing,these proceedings).In addition,detailed astrometric monitoring of evolved stars could reveal information about the existence of binary components,or even planets in these systems.Moreover,by monitoring its position new facts can be learned about maser physics.Are spots persistent,what (turbulent)motions can be measured,is the brightness a function of stellar cycle?Finally it is worth mentioning that the maser astrometry could tie the observational reference frames,in particular in the infrared where these stars are very bright.Carrying out astrometric monitoring projects with spectral line VLBI is extensive work that requires some special techniques and assumptions. There are a number of technical advancements and initiatives that will make improvements to thisfield which will make it much more accessible.1.Methods&TechniquesThe most straightforward way of doing VLBI astrometry is to use “phase referencing”to a bright and close calibrator in order to overcome the spatial and temporalfluctuations of the ionosphere and/or tropo-sphere.The calibrator is usually an extragalactic source with afixed and known position,which is sufficiently bright to calibrate the atmo-spheric and instrumental effects within a fraction of the coherence time. Because the modern VLBI correlators have incorporated the best geode-tic models,phase referencing results obtained in this way can be accurate to a few mas.This is already sufficient for a number of applications.In other cases,higher accuracy is required and some adjustments are nec-essary to improve the model.In such cases one probably needs to return to the”totals”in order to use astrometric/geodetic software.To get accuracies below1mas is not trivial;better models of the atmospheric behavior and structure of the reference source are required.For maser astrometry an additional complication can be that one needs to observe in mixed bandwidth mode,applying a narrow band for the maser detection and simultaneous wide bands for the referenceAstrometry of circumstellar masers3 source.Obviously the maser dictates the observing frequency,which rules out some calibration schemes used in continuum and geodetic VLBI.Ideally the masers should be bright and persistent.The brightness relates to the size of individual maser features,which sets an upper limit to the resolution one can obtain.A fundamental issue remains to link the motion of the maser spot to the stellar properties.One way forward is to assume that the masers are on a linear path with respect to the star (e.g.a constant outflow).In this case the parallax of individual spots equals the parallax of the star,and the average motion may be assumed to be the stellar motion.In other cases one may be able to deduce the position of the star from the distribution of the masers,for example when they form a ring.If we can determine the position of the star,one can in principle measure the motion and the parallax.Note that only when one can consistently determine the stellar position with(much)better than1AU accuracy,a useful parallax can be determined.In some cases there is evidence that special maser spots exists that correspond to the stellar continuum amplified by the maser shell.Such spots are then tied very accurately to the stellar position.In other cases one has to worry about the systematic and turbulent motions in the maser shell. In any case,there may be additional motions involved,for example when masers occur in binary stars.2.OH MasersThe OH masers usually originate in large shells at1000AU from the mass-losing stars.As the outflow is constant at that point,the maser radiates most effectively radially,from the front and back cones of the shell.Three18cm transitions can give reasonably bright maser emission.However,the VLBI spots are usually resolved at20mas, sometimes degraded further by angular broadening in the interstellar medium,notably in the direction of the Galactic centre(Van Langevelde et al.1992).Although the shells are large,the“amplified stellar image”paradigm makes the OH maser interesting for astrometry.Radio continuum,am-plified by the maser results in a spot that isfixed on the stellar position, bright and persistent.The observation that the most blue-shifted spots in two OH transitions coincides is a very good confirmation of this idea (Sivagnanam1990).Recently,comparison of a proper motion study of U Her with Hippar-cos agreed to15mas(van Langevelde et al.2000),thus confirming that the most blue-shifted and most persistent spot indeed follows the stellar4Figure1.Proper motion of the most blue-shifted spot in the1667MHz maser of U Her.A motion ofµvlbi=−15.57±0.56,−9.66±0.61mas/yr and a parallax ofπvlbi=3.85±1.14mas arefitted to the data.For comparison the position of a secondary reference source is also displayed.trajectory in this case.Therefore the maser motion can be used to ob-tain the parallax and proper motion of the star.U Her was observed10 times in6years on the NRAO VLBA which resulted in thefirst VLBI parallax of a mass-losing star(Fig1;Van Langevelde et al.2000).The motion agrees within the error with the Hipparcos data,from which no reliable parallax could be obtained.If the amplified stellar image is a general feature in circumstellar OH masers,it will allow astrometric measurements of a large number of stars. However,the modest brightness of OH masers limits the application of this technique to stars closer than1kpc.3.H2O masersThe22GHz masers are much brighter than the OH masers in circum-stellar shells.Therefore positional accuracies in the order of50µas are in principle achievable.Of course at this high frequency the coherenceAstrometry of circumstellar masers5 time is much shorter and phase referencing VLBI at22GHz is quite a bit more challenging.Additional complications arise from the fact that the circumstellar H2O masers are rather messy.The structures are often ring-like on scales of100AU,as expected from tangential amplification.However, large departures have been observed,as well as considerable variability. If the structures are persistent enough one can attempt to track individ-ual maser spots over a year and measure the parallax of the star from these(e.g.Kuruyama et al.,these proceedings).If this approach works the water maser observations offer the interesting possibility to measure parallaxes throughout a large fraction of the Galaxy.There have been claims that in water masers the amplified stellar im-age can also occur(Marvel1996,Colomer et al.2000).To check this we have observed U Her at22GHz with MERLIN,which yields a com-parable resolution as our previous OH VLBI ing the same position calibrators wefind,somewhat surprisingly,that the brightest H2O maser coincides with the optical position and the OH image.Al-though the MERLIN observations do not resolve individual maser spots observable with VLBI,we take this as a clue that amplification of back-ground radiation may play a role in water masers too(Fig2;Vlemmings et al.2002).4.The use of SiO masersThe SiO masers around evolved stars can be even brighter.At43GHz the spots sizes allow10µas accurate positions.At the same time the technical difficulties are considerable.The coherence time is only of the order of a minute and in addition there are fewer suitable extragalactic calibrators.The masers themselves are located close to the star,within10AU. They are bright and abundant.Monitoring of their structure indicates that the masers are showing motions on the time-scale of the variability of the star(Diamond&Kemball1999).Often these motions are not linear and it will be a challenge to measure a parallax from the combined motions of SiO masers.As the masers are in a ring close to the star, it will be possible to estimate the stellar position from the distribution of SiO masers.The proper motion of the star can be measured this way,even if the individual spots cannot be resolved.Proper motions of individual stars can probably be measured throughout the Galaxy using the43GHz transition.A very suitable target for stellar astrometry by SiO masers are the Galactic center stars.In this region there is an abundance of maser stars6Center at RA 16 25 47.468 DEC 18 53 32.849Peak flux = 1.9579E+06 JY/B*HZ Levs = 1.958E+05 * (1, 3, 5, 7, 9)M i l l i A R C S E C MilliARC SEC 100500-50-100806040200-20-40-60-80-100Figure 2.The location of the U Her H 2O masers with respect to the stellar positions determined from the Hipparcos observations,as determined from phase referencing to an extragalactic calibrator.The star is denoted by two circles indicating the star itself and the stellar radio photo-sphere.The error bars on the maser feature are the errors due to the positional fitting.The errors on the stellar position are due to the link to the radio-reference frame and due to the errors in the proper motion used to transpose the optical position.The triangle denotes the stellar position when using the phase referencing results on the secondary calibrator.Astrometry of circumstellar masers7 close together(Sjouwerman,these proceedings).In addition,SgrA*is a suitable calibrator;it is bright at short wavelengths and,as it is is the location of a massive black hole,it is stationary.In fact,an important results from circumstellar maser astrometry is the tie between the IR and radio positions in the Galactic centre.By locking the positions of the IR sources to the radio maser counterparts,it can be shown that the motion of stars in the Galactic centre is consistent with motion around a massive black hole located at the position of SgrA*(Menten et al.1997; Eckart et al.2002).Several groups have explored the possibility to start monitoring SiO masers in the Galactic centre region with VLBI(Sjouwerman et al.1998; Deguchi,these proceedings;Imai,these proceedings;Reid,private com-munication).The basic goal is to add two more components to the mea-surement of kinematics in this dynamically interesting region(Winnberg, these proceedings).5.FutureThere are many planned improvements of ordinary VLBI that will facilitate more astrometry of circumstellar masers.Several modern,big telescopes are under construction that will observe at high frequencies, necessary for H2O and SiO masers.Disk-based recording systems will improve the reliability and productivity of VLBI,and will be especially advantageous for time consuming astrometric monitoring.In addition there are several indications that the atmospheric modeling can be im-proved considerably.Other improvements of the geodetic models are important too,for example accurate telescope positions.However,the future for astrometry of masers clearly lies in instru-mental setups that overcome the temporal variability of the atmosphere at high frequencies.Some of this can be done with“cluster–cluster”VLBI,but the dedicated VERA project provides a far more ambitious approach that will soon produce interesting results(Kobayashi&Sasao, these proceedings).A number of studies are required to establish new techniques for as-trometry of circumstellar masers.Careful analysis of errors and con-sistency checks will be necessary to determine how to use the masers optimally to determine the underlying stellar motions.But in potential this method enables the measurement of proper motions and distances of mass-losing stars throughout the Galaxy.This will progress the study of stellar evolution as well as our knowledge of the structure of the Galaxy.86.AcknowledgementsThe National Radio Astronomy Observatory is a facility of the Na-tional Science Foundation operated under cooperative agreement by As-sociated Universities,Inc.MERLIN is a National Facility operated by the University of Manchester at Jodrell Bank Observatory on behalf of PPARC.We acknowledge the continuous support by Phil Diamond, Harm Habing and Richard Schilizzi.ReferencesColomer F.,Reid M.J.,Menten K.M.,Bujarrabal V.,2000,AA355979Diamond P.J.,Kemball A.J.,1999,in“Asymptotic Giant Branch Stars,IAU Sympo-sium#191”,eds Le Bertre,Lebre&Waelkens,p.195Eckart A.,Genzel R.,Ott T.,Sch¨o del R.,2002,MNRAS331917Marvel,K.B.,1996,Ph.D.thesis,New Mexico State UnivMenten K.M.,Reid M.J.,Eckart A.,Genzel R.,1997,ApJ475L111Sivagnanam P.,Diamond P.J.,Le Squeren A.M.,Biraud F.,1990,AA229171 Sjouwerman L.O.,van Langevelde H.J.,Diamond H.J.,1998,AA339897Van Langevelde,H.J.,Frail,D.A.,Cordes,J.M.,Diamond,P.J.,1992,ApJ396686 Van Langevelde H.J.,Vlemmings W.,Diamond P.J.,Baudry A.,Beasley A.J.,2000, AA357,945Vlemmings W.H.T.,Van Langevelde H.J.,Diamond P.J.,2002,AA in press。